1. Field of the Invention
The invention relates to mobile communication devices, and in particular, to intention based communications for mobile communication devices.
2. Introduction
Mobile communication devices such as cellular telephones have been becoming increasingly pervasive in our daily life. Significant progress has been made in providing a voice dialogue system for the mobile devices in an attempt to make them easier to use. To achieve this goal, the instruction or dialogue must be made easy to learn, easy to remember, and easy to retrieve.
To enable a user's voice to control a mobile communication device, conventional practice is to establish communication parameters between user and the cell phone. User has to know in advance the commands the cell phone can understand in terms of the dialogue system.
In addition, users desire the ability to send a short message to another user. However, this is a very difficult problem because that dictation process would consume a large amount of processing resources and out-of-vocabulary words cannot be handled. In this manner, it would be preferable to use a language model in the dictation mode, but to train such a language model is challenge, where data is scarce and dynamical.
On an object-oriented level of abstraction, interaction between objects can be seen as a form of communication. The goal of this communication is to change the state of two objects after messaging. The ways of sending messages are overloaded functions or methods. That means the above mentioned communications can be realized by transferring meanings or intentions rather than passing through exact words.
However, to implement a voice dialogue system on the mobile communication devices, we have to overcome the challenge of limited CPU power and memory resources. The trade off is often achieved by restricting the recognition grammars so that users have to follow a limited set of prescribed commands. Otherwise, the system cannot work. On the other hand, a voice dialogue system implemented on the server side can have more flexibility to allow a user to choose words.